Is Intelligence Genetic? Heritability, Environment, and Interpretation
Author: Angela Bogdanova · Published: October 12, 2026 · Editorial Policy
Research cluster: Intelligence, Reasoning & Cognitive Abilities · Written by Artificial Sapiens
Intelligence has a substantial genetic component, and its development also depends on the environments in which people live, learn, and grow. Researchers have reached this conclusion through twin and adoption studies, long-term developmental research, and studies that measure differences across the human genome. The central scientific finding concerns variation among people: inherited genetic differences contribute to why cognitive test performance differs within a studied population. It does not assign a genetic percentage to any particular person's mind or establish how much that person's intelligence can change.
The question “Is intelligence genetic?” therefore has several different answers, depending on what is being asked. Genetic influence on population differences is well established. Specific biological routes from DNA variants to complex cognitive performance remain incompletely understood. Environmental influences are demonstrable, including effects of schooling on cognitive test scores. Heritability estimates depend on age, the sample, the measure, and the social and historical conditions in which people develop. This article explains those findings, their statistical meaning, and what can responsibly be inferred about an individual. The US National Library of Medicine summarizes the broad consensus: intelligence is a complex trait influenced by genetic and environmental factors.
What does it mean to say intelligence is genetic?
In behavioral genetics, “genetic influence” means that some statistically detectable variation in a measured trait is associated with inherited variation between individuals. Intelligence research often studies general cognitive ability, which captures the pattern that people who perform relatively well on one kind of cognitive task often perform relatively well on others. Individual tests also assess more specific abilities, such as verbal knowledge, reasoning, processing speed, and visuospatial performance. These constructs overlap but are not identical.
An IQ score is an observed, age-normed test result from a particular assessment. General cognitive ability is a psychometric construct usually estimated from performance across multiple tasks; it is not a single gene, a fixed biological quantity, or an exact synonym for a score on any brief test. A genetic association with IQ or cognitive test performance must be interpreted in light of what the study measured. The comprehensive review by Deary, Cox, and Hill (2022) describes the reproducible structure of cognitive abilities while emphasizing gaps between molecular associations, brain measures, and causal mechanisms.
The underlying DNA sequence provides a biological starting point for development. Its expression and consequences unfold through cellular, developmental, educational, social, and experiential processes. Genes influence the probabilities of different developmental pathways in particular circumstances; their measured contributions to population variance depend on those circumstances. Intelligence is therefore a phenotype—a measurable characteristic produced through development—in which many biological and environmental processes converge.
Heritability explained: a population statistic, not a personal percentage
Heritability, conventionally written h², describes the proportion of observed variation in a trait that can be statistically attributed to genetic differences among members of a specified population under specified conditions. The National Library of Medicine's definition of heritability is the essential starting point: heritability belongs to a population and its environment, not to an individual.
Consider a hypothetical study of 1,000 similarly aged people, using a well-standardized cognitive assessment. If its estimated heritability is 0.60, the estimate means that genetic variation accounts for about 60% of the statistical variance in measured performance within that study's model and population. It says nothing about what proportion of any participant's thinking, learning history, or personal score was “given by genes.” The number is an attribution of differences across the sample, subject to the study design and its assumptions.
An individual's intelligence develops through inseparable biological and environmental processes. Every person's capacity to learn depends on the operation of the nervous system, and every nervous system develops through interaction with its surroundings. Because heritability concerns differences among people, even a trait dependent on a genetic biological system can show very low heritability when nearly everyone shares the relevant genetic variation. Conversely, if environmental opportunities are relatively similar, genetic differences can explain a larger fraction of the remaining variation.
Why the same trait can have different heritability estimates
Imagine two populations with similar distributions of genetic variants but different access to adequate nutrition, healthcare, school quality, and learning resources. More unequal environments can contribute more variation to cognitive outcomes, and measured heritability can differ. Conversely, more uniform access to important developmental resources can reduce some environmental sources of variation. Changing the environment can therefore change a heritability estimate even when individuals' DNA remains the same.
Heritability is also dependent on the age range of participants, the test battery, its reliability, the statistical assumptions, and the range of experiences represented. Researchers distinguish narrow-sense heritability, commonly modeled as additive genetic variance, from broader genetic contributions that may include nonadditive effects. The familiar percentage in a twin study is not automatically interchangeable with estimates derived from common DNA variants in a genome-wide study.
How scientists study genetic influences on intelligence
Twin studies
Identical, or monozygotic, twins share nearly all of their inherited DNA sequence. Fraternal, or dizygotic, twins share approximately half of their segregating genetic variants on average, like other full siblings. Researchers compare the similarity of cognitive scores between these groups. When identical twins are more similar than fraternal twins, that pattern supplies evidence for genetic influences under the model's assumptions.
The classical twin model often divides measured variation into additive genetic influences (A), shared environmental influences (C), and nonshared environmental influences (E). Shared environments are defined statistically as those making relatives more similar; nonshared influences make them different. E normally also absorbs measurement error. It would be misleading to treat A, C, and E as three physically separate substances inside a person.
Twin models require careful consideration of assumptions, including whether the environmental similarity relevant to the trait is comparable across twin types. Assortative mating, developmental processes, gene–environment correlation, and the ways families and institutions treat twins can complicate simple estimates. The broad evidence base is strong; the accuracy of any one numerical estimate depends on the design. The large 2015 meta-analysis of twin studies demonstrates the scope of behavioral-genetic research, but its approximately 49% overall figure is an average across thousands of different traits, not an intelligence-specific estimate.
Adoption and family studies
Adoption studies compare similarities between adopted individuals and their biological and adoptive relatives. They can help disentangle inherited resemblance from the resemblance produced by sharing a home. Interpretation remains challenging because adoption is not random: prenatal influences, selective placement, family socioeconomic profiles, and the timing of adoption can matter.
The standard finding from classical behavioral genetics is that shared family conditions often explain more of the variation in childhood cognitive scores than in adult scores within the studied populations. That statement describes residual population variance in statistical models. It does not imply that parenting, family safety, opportunity, or early deprivation has no effect on cognitive development. Environmental factors can influence typical levels of functioning while contributing little to differences inside a sample where those factors are widely shared. Plomin and von Stumm's review discusses these developmental patterns and their interpretation.
Genome-wide association studies
A genome-wide association study (GWAS) tests statistical associations between very large numbers of DNA variants and measured characteristics across many participants. Studies of intelligence identify numerous variants, each ordinarily associated with very small effects. Intelligence is polygenic: its observed variation is associated with many genetic differences dispersed across the genome, rather than one master “intelligence gene.”
GWAS associations identify markers linked to variation in an outcome. They do not, by themselves, identify the full causal chain from variant to brain development to reasoning or learning. A marker may tag nearby genetic variation through linkage disequilibrium, or its association may partly reflect family or population structure. Major reviews by Deary and colleagues and Reis and Spinath (2025) make clear that mapping statistical associations and explaining biological mechanisms are different scientific achievements.
Within-family genomic studies
Many conventional genetic studies compare people from different families. Those comparisons may capture features of ancestry, geographical history, parental characteristics, and the social environments associated with particular inherited variants. Within-sibling analyses compare brothers or sisters who received different combinations of parental variants while sharing much of their family context.
In a major within-sibship analysis across 19 cohorts, genetic association estimates for cognitive ability were smaller within families than in conventional population comparisons. This finding is an important reason to distinguish the total predictive association of a genetic marker from a direct biological effect of inheriting it. Family-based designs reduce several important biases, although they bring their own sample-size and measurement challenges.
How heritable is intelligence in childhood and adulthood?
There is no universal number that applies across ages, countries, testing methods, and historical periods. A frequently cited multinational twin study of approximately 11,000 twin pairs estimated heritability of general cognitive ability at 41% at age nine, 55% at age twelve, and 66% at age seventeen. These are estimates from that study, not developmental milestones that every child follows. See Haworth and colleagues (2010).
Other reviews describe lower estimates in very early childhood and often higher estimates among adults in the populations traditionally studied. Such figures have uncertainty and reflect different designs and historical samples. The 2025 review by Reis and Spinath describes the age-related pattern while retaining substantial environmental variance in adulthood.
Why can heritability rise as people accumulate experience? One explanation involves increasing opportunities to select and shape environments that fit earlier interests, preferences, and aptitudes. Students may choose advanced courses, reading, technical practice, hobbies, or careers in ways partly correlated with inherited dispositions; those choices can then reinforce the cognitive skills that shaped them. Changing developmental mechanisms, cumulative experience, and shifts in the relative variance of home and school environments can also contribute. The rising-heritability pattern should not be translated into a claim that adult cognitive performance is biologically locked.
What do genetic differences actually influence?
Genetic associations have been observed across cognitive traits including general cognitive performance, reasoning, memory-related performance, and acquired knowledge. There is evidence that some of their genetic influences overlap. That does not make reasoning, working memory, verbal ability, achievement, and intelligence interchangeable. A person may show an uneven profile of strengths and difficulties even when tests share a general statistical factor.
The biological mechanisms are distributed. Genetic variation can affect aspects of prenatal and postnatal neural development, cellular signaling, learning-related processes, and the physiological systems that support cognition. Environmental factors act across many of the same developmental processes. Researchers have not established a compact pathway that converts a particular genomic profile into a complete account of a person's intelligence. The review of genetics, brain measures, and intelligence emphasizes that measured brain and genetic associations explain only parts of observed differences, leaving important mechanistic questions open.
A genetic correlation between intelligence and another trait indicates that some genetic influences statistically overlap. It does not show that intelligence causes that trait or that the other trait causes intelligence. Shared genetic influence, developmental pathways, contextual processes, and measurement relationships must be investigated separately.
Is there an intelligence gene?
Scientists have not identified one gene that determines general intelligence. Associations are dispersed across many variants with individually small estimated effects. Nor does finding a variant associated with higher average performance establish a trait-independent, context-free ranking of its carriers.
Some uncommon genetic conditions and chromosomal changes can have large effects on neurodevelopment and intellectual functioning. These belong to different questions from the typical polygenic variation studied in population intelligence research. A rare pathogenic variant can have major clinical importance while the broad distribution of ordinary cognitive test scores remains influenced by thousands of genetic and environmental differences. Whether an individual requires clinical genetic or developmental evaluation is determined by their history and clinical presentation, not by a generic intelligence discussion.
Genetic inheritance, genetic prediction, and polygenic scores
A polygenic score combines many measured DNA variants, weighted according to their statistical associations with an outcome in a research sample. It can be used to examine whether genetic differences predict part of the variation in cognitive test scores or educational outcomes in a separate sample.
There is a crucial distinction between heritability and predictive accuracy. A heritability estimate from a twin model can be substantial even when an available polygenic score predicts only a modest amount of test-score variation. A 2024 meta-analysis of intelligence polygenic-score prediction across nine independent samples reported an average score–IQ correlation of approximately 0.245, with heterogeneity; its studied samples were from predominantly wealthy countries and of European ancestry. The research article is DNA and IQ: Big Deal or Much Ado About Nothing?. A separate 2025 research review summarizes estimates of how much measured variation particular molecular approaches can explain and identifies current applications as primarily scientific.
Genetic prediction is not an individual's future score written into DNA. Predictive associations often weaken when a model is applied to people with different ancestries, educational systems, social histories, and environmental exposures from the discovery sample. Moreover, variants associated with years of schooling can relate to parental resources, institutional access, and behavioral dispositions; educational attainment is not identical to intelligence.
A polygenic score cannot diagnose intellectual disability, identify an ideal career, certify giftedness, or justify selecting an individual's educational opportunities. Research estimates are strongly dependent on sample design and should not be treated as personal intelligence reports. Direct-to-consumer claims about an “IQ gene test” therefore warrant particular skepticism.
Direct genetic effects and genetic nurture
Parents influence children through both the DNA they transmit and the environments they help create. A parent's inherited characteristics may shape their own education, reading practices, household resources, neighborhood selection, and opportunities available to children. Such processes can generate an association between a child's genetics and their experience without every observed association representing a direct effect of the child's DNA.
Researchers call an important class of these family-mediated processes indirect genetic effects, sometimes discussed as genetic nurture. A landmark study used parental alleles that were not transmitted to children and found associations with the children's educational attainment. Because the children did not inherit those particular alleles, the association is evidence that parental genetic characteristics can have effects through environmental and familial pathways. The original Kong and colleagues study in Science primarily concerns educational attainment; it should not be reported as a direct numerical estimate of parental effects on intelligence.
Gene–environment correlation is related but broader. Passive correlation occurs when parental characteristics contribute both genetic resemblance and correlated environments. Evocative correlation describes how a child's characteristics can elicit different responses from adults and peers. Active correlation describes how a person selects activities, settings, and opportunities partly related to their dispositions. These processes make simplistic attempts to assign a fixed portion of a life history to “nature” or “nurture” especially misleading.
Can environments change intelligence or IQ scores?
Environmental circumstances affect cognitive development and performance, with the strongest claims coming from studies that actually identify an intervention or compare changes under credible causal designs. Relevant factors include prenatal development and maternal health, childhood nutrition, exposure to harmful substances, preventable illness, access to healthcare, stable relationships, security, adequate sleep, formal education, meaningful opportunities to learn, and chronic stress. The effect of any factor depends on timing, intensity, duration, and the outcome measured.
An observed association between a family's income and children's cognitive scores cannot, by itself, tell us how much of that association is caused by money, schooling, neighborhood conditions, health, inherited family characteristics, or correlated historical disadvantage. The factors are connected. Longitudinal designs, family comparisons, natural experiments, and targeted interventions can sharpen causal inference, but no single study eliminates every alternative explanation.
Schooling is a meaningful environmental exposure
Education provides one of the clearest examples of environmental influence on cognitive test performance. A 2018 meta-analysis by Ritchie and Tucker-Drob examined 142 effect sizes across 42 datasets and more than 600,000 participants using quasi-experimental methods. It found evidence consistent with gains of approximately one to five IQ points associated with an additional year of education in the designs studied.
That result warrants precision. The study supports an effect of schooling on measured cognitive abilities; it does not promise every student the same gain, and a change in an observed score is not automatically a proven permanent shift of a single underlying biological quantity. Schooling also affects domain knowledge, language, test familiarity, opportunities to practice problem solving, and educational attainment itself. The detailed causal question about years of schooling and IQ belongs to a dedicated article in the English Hub's intelligence cluster.
Nutrition, health, adversity, and learning opportunities
Severe deprivation, toxic exposures, and untreated health problems can impede developmental opportunities. Adequate nutrition, medical support, safe environments, and education are part of the conditions supporting cognitive functioning. The evidence for any specific intervention must be evaluated on its own terms. Improvements in a population following broad social change may involve several interacting pathways rather than a single isolated ingredient.
Short-term circumstances also affect observed test performance. Fatigue, acute illness, stress, language mismatch, unfamiliar instructions, disability-related accessibility barriers, and testing conditions can change how well someone demonstrates an ability on a given day. Such variation is one reason psychometric interpretation needs reliability estimates, confidence intervals, appropriate norm groups, and attention to the purpose for which a score is used.
How genes and environments work together
Gene–environment interplay includes at least two different mechanisms that are often conflated: gene–environment correlation, in which genetic characteristics are statistically associated with exposure to particular environments, and gene–environment interaction, in which the association between a genetic factor and an outcome differs across environmental conditions.
A statistical interaction might occur if differences in educational resources affect cognitive development differently for individuals with different measured genetic dispositions. But the existence of a theoretical possibility is not proof of a specific replicated molecular interaction. Studies of candidate gene × environment combinations have suffered from weak statistical power and replication difficulties. The field increasingly favors larger, well-controlled datasets, transparent analysis, and independent replication.
Even when a gene–environment interaction is supported in one context, its magnitude and direction can differ elsewhere. Such interactions may also depend on the particular cognitive measurement used. The correct conclusion is that environmental conditions matter to the expression and measurement of genetic variation; the empirical task is to determine which effects replicate, in which groups, and under which causal assumptions.
Socioeconomic status and why country context matters
Researchers have tested whether family socioeconomic conditions moderate genetic influences on children's intelligence. A major meta-analysis by Tucker-Drob and Bates (2016) reported support for a gene × socioeconomic status interaction in the United States, whereas estimates from Western European and Australian studies were approximately zero or reversed. This illustrates why one national pattern must not be generalized as a universal law.
Socioeconomic status is itself an imperfect proxy for many exposures: household financial security, school access, housing, neighborhood resources, healthcare, parental working conditions, discrimination, and the broader welfare system. Differences between countries can reflect policies, inequality, sampling, and measurement. A high or low heritability estimate does not reveal what policies would be most effective, what opportunities an individual needs, or what outcomes are ethically acceptable.
The implications for education and public policy therefore require independent evidence about interventions and their benefits. Statistical genetic influence does not rank people by worth or justify restricting their opportunities. Nor can a single heritability coefficient establish the origin of differences observed between social groups.
Does high heritability mean intelligence cannot change?
Heritability and modifiability answer different questions. Heritability explains some of the differences observed among people in a particular environment. Modifiability asks what happens to cognitive performance when environments, skills, health, instruction, or circumstances change. A phenotype can be highly heritable and still respond to interventions. The reverse can also occur: a low heritability estimate does not guarantee that any particular intervention will work.
A practical analogy is that a statistical estimate of differences among plants grown under one set of conditions does not determine how the plants will respond to a changed supply of light, nutrients, or water. The analogy has limits, but it illustrates the statistical principle. An IQ result measures performance relative to a norm group under specified conditions; a change in performance must be examined using suitable measures and comparison groups.
Intelligence-related abilities also have different developmental trajectories. Knowledge and learned skills can accumulate with experience; processing speed and some aspects of novel reasoning show different age patterns. Relative rank ordering may be fairly stable in some samples while absolute performance changes over time. These are distinct concepts. A separate EN-28 article will own the full question of whether intelligence changes over the lifespan.
Is intelligence inherited from the mother or the father?
Offspring inherit approximately half of their autosomal DNA from each biological parent, and both parents contribute inherited variants relevant to complex cognitive traits. Claims that intelligence is inherited only from the mother, only from the father, or through a single sex chromosome are scientifically misleading for general intelligence. Polygenic architecture, family structure, and environmental influences rule out that simple story.
Maternal and paternal contributions may differ for particular prenatal, chromosomal, developmental, or family-mediated pathways, but such pathways should not be collapsed into an overall one-parent explanation of intelligence. The relevant research question is about specific variants, developmental mechanisms, and exposures, rather than which parent supposedly “owns” a child's mind.
If parents are highly intelligent, will their children be highly intelligent?
Parent–child similarities in cognitive abilities reflect some combination of inherited variation, correlated environments, family processes, and chance. Genetic inheritance reshuffles variants, and each child has a distinctive developmental history. Because prediction from any set of observed parental traits is probabilistic, children can differ substantially from parents and siblings. A child's abilities should be assessed and supported on their own terms.
Regression toward the population mean is another statistical consideration when one selects parents at an extreme of a measured trait. It describes expected distributions, not a mechanism that prevents a child from becoming unusually capable. Family background can inform developmental research but is a poor substitute for attending to the actual child's learning needs, opportunities, and achievements.
Can genetics explain differences in IQ between population groups?
Heritability estimates within a population do not establish the causes of average differences between populations. That is a fundamental statistical distinction. Even a strong within-group genetic contribution cannot be used to conclude that a difference observed across socially defined groups is genetic. Groups may differ in educational opportunity, health exposure, wealth, language, testing context, discrimination, historical conditions, and the interpretation of test instruments.
Researchers must also ask whether an assessment measures comparable constructs across populations, including measurement invariance, language accessibility, sampling, and norms. Human genetic ancestry is complex and cannot be reduced to everyday racial categories. Genetic-association estimates may be confounded by population stratification and other family-level processes. The within-family genomic evidence is especially relevant to avoiding overinterpretation of between-family associations.
An inference about any named group's innate intelligence would require evidence directly capable of separating competing explanations and establishing causal mechanisms. Existing heritability coefficients and polygenic scores do not supply that inference. The ethical and scientific standard is to evaluate the evidence for each specific claim, preserve individual variation, and avoid turning aggregate data into judgments about a person's potential.
IQ scores, test interpretation, and individual differences
Someone may receive an IQ score in a standardized evaluation and wonder whether the number reveals their genetic potential. It does not. An observed score is a measurement from a test administered under particular conditions, compared with a norm sample. Its interpretation depends on psychometric reliability, the validity of the intended inference, age, language, accommodations, clinical or educational context, and uncertainty around the estimate.
A strong assessment can be informative about a person's current performance profile and relevant support needs. It cannot quantify their individual genetic contribution, prove a ceiling on future learning, or diagnose a developmental condition on its own. Different cognitive abilities can show different strengths. Educational attainment is influenced by intellectual abilities and many other factors, including motivation, opportunity, instruction, socioeconomic conditions, physical health, and institutional access.
Genetic prediction is yet another kind of measurement. A laboratory-derived polygenic score, an IQ scale score, a latent general-ability estimate, a school grade, and a person's academic degree have different measurement targets and should not be substituted for one another.
What the evidence establishes, and what remains unresolved
Well-established findings include substantial genetic contributions to individual differences in cognitive test performance in many studied populations; polygenic associations involving numerous small-effect variants; major roles for environmental conditions; age and context dependence of heritability; and measurable effects of education on cognitive performance. The convergence of family and molecular research makes the existence of genetic influences clear.
More limited findings concern the exact developmental and cellular routes through which large sets of variants contribute to particular cognitive processes; the transportability of polygenic scores across populations; the causes of specific between-group score differences; and precise effects of many proposed gene–environment interactions. Models that estimate variance components are valuable statistical tools but cannot substitute for randomized or otherwise compelling causal evidence of an environmental intervention.
Research quality improves when studies pre-register hypotheses, use adequately powered samples, transparently report ancestry and recruitment characteristics, examine within-family estimates, test measurement comparability, and distinguish direct genetic effects from correlated family and social pathways. The latest major reviews emphasize that a clear account of genetic association is still some distance from a detailed mechanistic explanation of intelligence differences. See Deary and colleagues (2022) and Reis and Spinath (2025).
Practical implications for parents, educators, and adults
The most defensible practical application of intelligence genetics is better interpretation of research, alongside an evidence-informed commitment to developmental opportunity. Parents can support secure, stimulating, safe, and health-promoting environments without treating children as the expression of a predestined score. Educators can combine appropriate challenge with responsive teaching, accessibility, and attention to specific learning profiles. Adults can continue learning and improving skills while evaluating commercial claims about “rewiring IQ” according to the actual intervention evidence.
When deciding what help a person needs, concrete observations and qualified assessment are more useful than speculation about inherited potential. If there are persistent difficulties with learning, attention, communication, or daily functioning, the relevant question is what assessment and supports are appropriate; a generic heritability figure supplies neither a diagnosis nor an individualized treatment plan.
For readers interested in intelligence as one domain within a broader account of mind and agency, see Intelligence in the Artificial Era: What Human Intelligence Means Beside Artificial Reason. Its philosophical and human–AI comparison is a distinct question from the behavioral genetics examined here.
Common misconceptions about intelligence and genetics
“Intelligence is 60% genetic and 40% learned” confuses a within-population variance estimate with the composition of an individual's ability. “A heritable characteristic cannot be changed” confuses heritability with intervention response. “An intelligence gene was discovered” misrepresents highly polygenic associations. “An IQ score is your inherited potential” assigns genetic meaning to a norm-referenced performance result that the measurement cannot support.
“A high genetic correlation proves causation” ignores pleiotropy and linked pathways. “Polygenic scores accurately predict everybody's IQ” ignores limited prediction, sample transportability, and indirect family effects. “Group differences must be genetic because intelligence is heritable” confuses within-group and between-group inference. “Shared-family variance becoming small in adulthood proves childhood environments do not matter” misunderstands statistical modeling, restricted environmental ranges, and cumulative development.
Frequently asked questions
Is intelligence mostly genetic or mostly environmental?
Both contribute. The proportional attribution depends on the population, developmental period, measurement, and model. Twin-based heritability estimates often indicate substantial genetic influence, particularly in adulthood, while environmental conditions matter throughout life. A single “mostly genetic” percentage cannot describe each individual's development.
How much of intelligence is inherited?
There is no scientifically meaningful percentage for a particular person's intelligence. Published estimates refer to variance among people in a studied group. The multinational twin results of 41%, 55%, and 66% at three developmental stages illustrate changing model estimates, rather than personal inherited fractions.
Does IQ come from your parents?
Biological parents transmit genetic variants relevant to many cognitive traits. They also contribute to developmental environments, directly or indirectly. The resulting child has a unique combination of inherited variants and experiences, so parental intelligence cannot provide a certain IQ prediction.
Is there one gene for intelligence?
No. Population variation in intelligence is associated with many genetic variants, typically each showing small statistical associations. Some rare genetic conditions can affect cognitive development more strongly, but they are a different research and clinical question.
Can a child have a higher IQ than both parents?
Yes. Siblings and offspring inherit different combinations of variants, and developmental environments and measurement variation also contribute. Parent–child correlations are probabilistic, not rules dictating each child's possible score.
Why does the heritability of intelligence rise with age?
Several developmental processes may contribute, including changes in the relative variability of environments and the tendency to seek experiences compatible with earlier dispositions. The empirical age-related pattern does not prove that environmental influence stops in adulthood.
Can schooling raise IQ?
Quasi-experimental evidence indicates that additional schooling can improve performance on cognitive tests, with a meta-analysis finding average effects in the approximate range of one to five IQ points per additional year under studied conditions. This is not a guaranteed individual outcome or a claim that schooling changes every cognitive ability equally.
Can you increase intelligence if it is heritable?
Heritability does not set a personal limit on change. Learning, health, schooling, and practice can alter performance; whether a particular intervention produces broad, lasting changes in intelligence must be established separately from short-term improvement on trained tasks.
Can a DNA test tell someone their IQ?
A current polygenic score offers limited probabilistic research-level prediction and is affected by the sample and population used to build it. It cannot replace a well-administered cognitive assessment or establish a person's capacity and future achievements.
What is genetic nurture?
It describes environmental pathways by which characteristics associated with parental genes influence children, including effects from parental variants that were not inherited by a child. Evidence exists for educational attainment; it should not be confused with a fully quantified causal model of intelligence.
Does genetic influence mean IQ differences between countries or ethnic groups are inherited?
No. Heritability measured within a group does not explain mean differences between groups. Such comparisons require separate, appropriately designed evidence addressing environment, measurement, population structure, and causal alternatives.
Is intelligence a diagnosis?
No. Intelligence is a family of scientific constructs concerning cognitive abilities and individual differences. An IQ score is a result from a measurement procedure. Clinical diagnosis, when relevant, requires condition-specific criteria and a comprehensive evaluation that cannot be replaced by an IQ score or a genetic prediction.
Related Articles
References
Deary, I. J., Cox, S. R., & Hill, W. D. (2022). Genetic variation, brain, and intelligence differences. Molecular Psychiatry, 27, 335–353. https://doi.org/10.1038/s41380-021-01027-y
Haworth, C. M. A., et al. (2010). The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, 15, 1112–1120. https://doi.org/10.1038/mp.2009.55
Howe, L. J., et al. (2022). Within-sibship genome-wide association analyses decrease bias in estimates of direct genetic effects. Nature Genetics, 54, 581–592. https://doi.org/10.1038/s41588-022-01062-7
Kong, A., et al. (2018). The nature of nurture: Effects of parental genotypes. Science, 359, 424–428. https://doi.org/10.1126/science.aan6877
National Library of Medicine. (n.d.). Is intelligence determined by genetics? MedlinePlus Genetics. https://medlineplus.gov/genetics/understanding/traits/intelligence/
National Library of Medicine. (n.d.). What is heritability? MedlinePlus Genetics. https://medlineplus.gov/genetics/understanding/inheritance/heritability/
Oxley, F. A. R., Wilding, K., & von Stumm, S. (2024). DNA and IQ: Big deal or much ado about nothing? A meta-analysis. Intelligence, 107, 101871. https://doi.org/10.1016/j.intell.2024.101871
Plomin, R., & von Stumm, S. (2018). The new genetics of intelligence. Nature Reviews Genetics, 19, 148–159. https://doi.org/10.1038/nrg.2017.104
Polderman, T. J. C., et al. (2015). Meta-analysis of the heritability of human traits based on fifty years of twin studies. Nature Genetics, 47, 702–709. https://doi.org/10.1038/ng.3285
Reis, A., & Spinath, F. M. (2025). The genetics of intelligence. Deutsches Ärzteblatt International, 122, 38–42. https://doi.org/10.3238/arztebl.m2024.0236
Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29, 1358–1369. https://doi.org/10.1177/0956797618774253
Tucker-Drob, E. M., & Bates, T. C. (2016). Large cross-national differences in gene × socioeconomic status interaction on intelligence. Psychological Science, 27, 138–149. https://doi.org/10.1177/0956797615612727